On this episode of TD Cowen’s Restaurant Rendezvous Podcast Series, Nuro’s Head of Partnerships Cosimo Leipold joins Andrew Charles, TD Cowen’s Restaurant Analyst to discuss Nuro’s mission and journey as the company’s robotic technology disrupts restaurant delivery.
Press play to listen to the podcast.
Welcome to Cowen Insights, a space that brings leading thinkers together to share insights and ideas shaping the world around us. Join us as we converse with the top minds who are influencing our global sectors.
Welcome to our next edition of Restaurant Rendezvous’s podcast. We’re pleased to be joined today by Cosimo Leipold, head of partnerships at Nuro. Cosimo, thanks so much for joining me today.
Thank you, and thank you for having me.
Cosimo, can we start off by giving us the quick 90 second elevator pitch on Nuro?
Absolutely. Maybe the easiest way to do that is, let me just start with what the three simple principles that the company was founded on are. Admittedly now, they probably seem a bit more obvious than they did in 2016, but bear with me for a moment. So they were as follows, right.
First, number one, that the cost of human delivery is going to get worse, not better, right. And this is obviously pre-COVID, pre-Great Resignation and all of those things, which further exacerbated the problems that exist in the industry already. The second one is on-demand was going to become the new normal, right. So what was five day shipping became two day shipping became same day shipping, and even now that’s not fast enough. And then finally there was this premise that good’s autonomy is a fundamentally different problem than passenger autonomy, and that it’s going to commercialize both more safely and more quickly than passenger autonomy.
Now there’s a lot I could say about that statement and I think that’s actually probably the most interesting thing I’ve learned in the last six years, but I’ll point to two things as proof points, if you will. Number one, we are the only company to my knowledge that has fully autonomous operations in three states. And second, we’re the first and only company to have received NHTSA exemptions for our vehicle and the first company to have received a deployment permit from the State of California.
And so I look at that as fundamentally a proof point of the technology and the approach that we’ve taken with goods and how that’s allowed us to become different than the rest of the industry. So where do we go from here, right? The next part of this is really about the mission that the company is on, and we believe very strongly we’re on a mission. We’re on a mission to better everyday life through robotics.
What that really means is that, not to get too sappy, but that we should be spending time on things that matter. Time with family, time with kids, time with friends. Times doing things we like to do, right. Not hours stuck in traffic or running errands that you would rather arguably avoid. But importantly, we want to build a service and a technology that enables that, not just for the lucky few frankly, but in a very equitable manner for the rest of the United States.
And so that’s what we’re now focused on, right. We’ve built this incredible technology. We’ve managed to operate autonomously in three different states. We’ve got all the regulatory approvals we could ever want. I mean, now it’s about scaling that service and preparing ourselves and our partners, frankly, for the cavalry that’s coming.
Cosimo, that’s great background on the mission of Nuro. Can you talk about what Nuro is building that to the untrained eye might look like a cool sidewalk rover?
So what we’re building I think is very different from the sidewalk robots, and in a couple of very key ways. Most notably, it’s an on-road vehicle, right. We’re classified as a low speed vehicle. From a regulatory standpoint we operate at speed, we operate on public roads, and we move 25 miles an hour today. We can go faster than 25 miles an hour. That’s a regulatory reason we chose to stay at 25.
But what that really means is that we’re able to carry much larger payloads because the vehicle is about half the width of a traditional sedan, it’s about three quarters of the length or so, carries about four or 500 pounds, has two large compartments. And each of those compartments can be further customized or segmented. We can create lockers and things of that nature. We also have the ability to do active cooling, active heating, things like that.
And so the product that was built was really built on this premise that, look, in the United States at least, the average delivery distances that you look at, primarily for restaurants but also frankly for other businesses, tend to be in that three to five mile range. And when you start looking at those kinds of distances you really need to be on-road to be delivering services and goods that deliver food, frankly, in a warm way or in a cold way, depending on what you’re delivering. If you’re doing that on a sidewalk, I think it’s much, much more difficult.
And the second part of that was we did analysis very early on, on the fundamentals of the sidewalk business versus us. And what we realized very quickly was that the sidewalk robots would have to proliferate to such an extent that they would very quickly go from being cute to being a nuisance. And our hypothesis was that people would not accept the appropriation of a public good for private gain, in this case sidewalks. But they would accept the idea that commercial activities happen on-road.
And I think that’s also proven to be true. There’s both instances in New York City where sidewalk robots had received cease-and-desist, I think from either the governor or something of that nature. And then in San Francisco, there’s actually been legislation introduced to limit the number of sidewalk robots. And so I think the strategy at its core around getting goods to people fast requires you to be on-road.
Maybe a different way to put it is this, the world has always lived on this efficiency frontier, if you will, right, for those in the audience that are economics geeks. And so you had to pick, right, between either you can be very fast and very expensive, and that’s the Instacarts and the Uber Eats of the world, or you can be relatively cheap and relatively slow, the traditional FedEx, UPS, things of that nature.
You can’t come off the curve, or at least you haven’t been able to come off the curve yet. What Autonomy allows you to do is to be both fast and cheap, but to do that you have to be on-road. And that’s one of the fundamental reasons that we built what we built as opposed to drones or sidewalk robots or some of these other technologies that are coming out.
Thanks for that, Cosimo. What is your role, leading partnerships at Nuro?
I was fortunate enough to join Nuro on day one. So my role has shifted over time. Originally I was, what I used to call one minus X, which was really the role of doing everything non-engineering, which by the way is good, because I’m not a good programmer, nor am I mechanically inclined. So it’s a good thing that I wasn’t involved in those things. But now my focus is much more on figuring out how to best lead Nuro, and frankly lead our partnerships towards shared success. And I really do mean the shared success part, right.
For a lot of other self-driving car companies, goods delivery is an add-on, a why not, might as well do this, right. But for us it’s our singular focus. So when we talk to partners about our commitment to helping their businesses, it’s a genuine discussion, right. It’s grounded in some ground truth and some real realities. There’s no competing priorities. Goods isn’t a secondary focus. It’s the only thing that we do. And that means that the partners that we have, which I’m happy to talk through, and the partners that we seek, really have this shared vision and we’re helping them build the right ecosystem to take advantage of autonomous delivery in the future, right.
One of the key questions for instance, that we try to answer for partners is, how well positioned are you for Autonomy? Are your stores in the right locations, and how do we know that they are or that they’re not? So frankly, should you be working with us? Should you be working with anybody? Or should you be thinking about something else entirely? And having that strategic thought partnership is a core part of how I see at least, leading partnerships at Nuro.
What are the commonalities for geographic areas where Nuro can find the most traction. And conversely, what are the similarities for geographies that’ll be most difficult to launch?
Great question. The way we think about it is as follows. We have two dimensions that we look at cities by. First, is the underlying dynamic of the city, right. The size, the opportunity, right. And the second has to do a lot with weather. So if you look at the places that we’re currently operating, Texas, Nevada, Arizona, there’s a common theme there, right. They tend to be sunny, fair weather locations.
Places with a lot of snow, so the Northern states by and large, are places we haven’t gone to yet, and places that we don’t feel, frankly, that we need to solve just yet. There’s an enormous amount of opportunity just in the Southern states, in the Sunbelt so to speak, where there’s already legislation in place in many of those states that are permissive for AV operation, like ours, and in many other places that it’s either coming or it’s planned. So lots of opportunities.
The other interesting thing about the way that we’ve approached it is, we didn’t just go to one state. We didn’t just go to Houston. We also went to Phoenix and we also went to Mountain. We also did all these things. And part of the reason we did that is we wanted to build a solution that was genuinely generalizable, right. Meaning, we don’t want to just solve Houston and then end up going to another state and realizing, oh, goodness gracious. Turns out the Autonomy system we solved is over-trained or over-fit to what we saw there and doesn’t work here.
And I think one of the things that we’ve been able to do by being in different cities, and arguably the cities that not a lot of other autonomous companies have gone to in some cases at least, is that we’ve been able to build a truly general purpose Autonomy system. So again, we operate fully autonomously with no safety driver in three different states today, and probably soon in Nevada, although we’re using it primarily as a test facility.
So it’s really about the scale of the opportunity in a state that we’re in. The regulatory environment, the weather, and then broadly, how we’ve approached the technology to ensure that it can be used everywhere when the time comes.
And then, Cosimo, maybe just the last question for background on Nuro. On a regulatory basis over the last year, what have you seen for AVs? Have more states or municipalities made it easier for you to expand into new markets?
Yes. And look, not to berate the point, right, but one of the things that I think has been very valuable for Nuro is the goods focus on a regulatory front. So consider this as an example, right. Every car you and I have ever bought protects us on the inside over what’s on the outside of the vehicle. And for good reason. Good luck selling a car that doesn’t do that, right.
But for Nuro, the absolute inverse of that is true, right. We can always protect what’s on the outside of the vehicle as opposed to what’s on the inside of the vehicle, right. I can replace eggs. I can replace pizza. I can’t replace people. So for example, one innovation that we have is on our third generation vehicle, we actually have an airbag on the outside of the vehicle. We’re literally taking safety innovations and flipping them inside out.
And the idea is that, look, if we ever get into an accident, right, we need to do everything we can to protect the pedestrians and the other road users. And the inside, we can always deal with later. And the benefits of this are really real, right. And conceptually it makes sense in its own right.
But we did a study with Virginia Tech Transportation Institute about a year ago to analyze the impact of zero occupant vehicles such as ours, right. So occupant-less vehicles. And what they found was that it reduces the risk of death or injury by about 60% for every mile of driving it replaces. So it’s a really significant safety opportunity, not just for us, but broadly for the industry. And I think regulators get that.
So to come back to your question for a second, there’s really two parts to regulations of AV vehicles, right. One is the federal level and the other is at the state level. Now at the federal level, the DOT finalized the rule this spring that makes it clear that vehicles such as ours, like Nuro’s, that never carry a person, driver or otherwise, don’t need to have occupant protection equipment, like seat belts and things of that nature.
So if you see our vehicles on the road, you’ll notice we don’t have a windshield, we don’t have a seatbelt obviously, we don’t have a steering wheel for that matter. We also don’t have little side view mirrors. We did on the first generation vehicle. We got the exception to remove those in the second generation vehicle. But the point here being that this zero occupant vehicle legislation is the first rule that the DOT has ever issued for AVs. And so I think they understand that goods delivery or goods autonomy is really the safest and best place to start dipping their toes into the legislation.
And on the state level, right, there’s three new states this year that introduced legislation. I believe those were Kansas, Oklahoma and West Virginia, right. That brings the total of the number of states to about 20, 21, 22, something around that order of magnitude, that now have permissive AV legislation. That’s about 55% of the American population is now covered. And the rest of the states are going to follow suit, right. Everybody wants AV technology. Nobody wants to be the state that doesn’t allow it.
And eventually, right, we’re going to get to a point where there’s a standard, uniform answer across the states and a standard, uniform answer for the federal government. But basically the trend that we’ve been seeing is a steadily improving landscape with very clear, and frankly, reasonable rules for deployment, and we think that’s just going to continue.
Super. I want to get into a few of the partnerships that you guys have active, and I believe Domino’s was the first major multi-unit restaurant partnership with Nuros. Can you talk about how that came to existence?
The way we met Domino’s was fortuitous and lucky, frankly, right. Coca-Cola holds an annual conference every year, or every other year, I’m not sure, where they have their largest customers attend. We came and we presented and Domino’s happened to be in the audience. We got very lucky, I think by that fact, because it turned out that we had a very common vision for the future, right. Domino’s knew and understood that it doesn’t make sense to use a Ford F150 truck for sake of argument, to move a pizza, right. You can use something much smaller and lighter and that’s much, much more efficient.
And so coming out of that conference we probably had Domino’s, I think maybe just even a week later they were in our offices, it was awfully quick. And within about 30 days or so, I think we had our first contract signed with them. So very, very quick from initial introduction to signature. And one of the reasons I think it’s gone so well with them, not just that we have this common vision, but they understand the complexities of this problem really well.
For instance, one of the things they realized was that you can’t rob Peter to pay Paul. Meaning, you can’t take savings from the delivery experience and then introduce friction in the store because it’s arguably possible that the friction in the store will offset whatever savings you have in the delivery space, right.
And so one of the things that we did together was we built a very integrated solution. So today if you get a delivery from a Nuro vehicle and it’s flowing through the store system, it’s fully integrated into the point of sale. Meaning, the people in the store, the store associates, don’t see the distinction until the very last minute, right.
It shows up on the same screens. They make the pizza the same way. It’s exactly the same process. There’s no additional friction. And at the very, very last step, there’s a moment at which the driver can see whether or not they’re taking the pizza or whether we’re taking the pizza. I mean, the instructions obviously are a little bit different.
But otherwise the intention was to make this as seamless as possible, all the way from the moment of order, all the way through the end customer experience at the moment of delivery. In fact, they also integrated everything into their app, right. So the things like being able to track the vehicle, knowing where it is, getting notifications about your pin codes and how to open the compartments, all of that happens natively within the Domino’s app. None of it is kluged together outside of that experience.
Thanks for that, Cosimo. And the other major concept you partner with is Chipotle, which is different because they took an equity stake in the Nuro business. Chipotle CFO last year indicated the use of Nuro by Chipotle has targeted the next five years even. Can you tell us more about what’s been disclosed about that partnership, including how many stores where it’s currently live?
Chipotle is an incredibly strong partner of Nuro, right. In addition to their stake in the company, we are collaborating with them on a number of different commercial opportunities, and unfortunately I can’t share an awful lot more than that. So I appreciate that’s not terribly satisfying as an answer. But more to come, stay tuned.
What I will say is this, Chipotle is an incredible partner and they share much of that same DNA that I think made others that we’ve worked with, like Domino’s, successful, right. They also recognize, and are frankly leaders in the digital space, and recognize the opportunity that is ahead of them, as well as how to make this experience really, really good. And what we’re doing now is we’re trying to take some of our learnings from other partners and make sure that we actually build the best possible mousetrap with them as well.
Cosimo, can we talk about your view about goods versus passengers? I mean, obviously what you guys satisfy is more on the goods side, but how do you guys think about it at Nuro?
Well, there’s really two big differences that I think make this unique. And in fact this was my very first question to Dave, one of the other co-founders, in the early days of Nuro, on the first day, actually. Look, how is this different, right? I mean, don’t you have to detect cats, dogs, pedestrians, and everything else in between. Aren’t we just solving the same problem all over again. And the answer is, no. It’s actually quite different.
And there’s two parts to that. First on the hardware side, right, because we never carry people. That means you don’t need stereos, leather seats, creature comforts. You don’t need as much space. That means you need less steel, less battery, less weight, which means the vehicles cost less. It means they’re safer, it means they’re more courteous. I think as I mentioned before, our vehicle’s about half the width of a traditional sedan.
So that means that we can do things, right. We can do things like pull over for somebody that’s tailgating and let them pass us very easily. We can park more easily, right. If we have to double park, we’re not blocking the entire lane. And then perhaps most importantly, as I mentioned before, we can design the vehicle to protect what’s on the outside, over what’s on the inside. So there’s some on the hardware side.
But the real magic, I think is on the software side, right. The people who are solving passenger autonomy have to solve both safety, right, obviously you have to be safe, and comfort. And comfort here means things like you don’t tap the brakes because maybe there’s something in front of you. You don’t pull over to the side of the road because there’s a cop doing interpretive dance to get people across the intersection. You don’t go the slower route around a neighborhood, right. You don’t do things like that because as a human you want to get between point A and point B as fast as possible. Effectively you want to be a more aggressive driver.
But with goods, suddenly all of those factors become malleable, right. Now we can do those things. We can pull over. We can wait for the interpretive dance cop to wave us across the road. We can take a longer route between point A and point B so long as we’re meeting the SLAs and the ETAs. And that means that we’re relaxing some of the hardest equations or some of the hardest parts of autonomy in a way that allows us to commercialize much, much more quickly, which is why we’ve been able to get out with all of these partners in the very short period of time that we’ve been in existence.
No, look, that’s very interesting. Maybe just two more questions, Cosimo. What are the characteristics of restaurants that you believe will succeed best utilizing Nuro?
So there’s a couple of things I’ve touched on already, right. So certainly a company culture that embraces innovation, right. We want folks that see the same future that we do. We want folks that are invested in mobile, invested in digital, that want to create a really amazing experience. We want this to be magical. So you need to have folks that share in that vision, but that’s obvious.
The second part is one that I actually think is really interesting and that we can help partners with, which is you want restaurants that are in the quote unquote, right location. Now, right is hard to define. But one of the things that we’ve we’ve built at Nuro is what we call a Bissum tool. What this allows us to do is allows us to take data from our partners.
So for instance, we can take every single location that somebody has across the United States and any historical delivery data they have, or parameters that approximate that data. So, number of deliveries per hour, estimated SLA, delivery radiuses, whatever we can simulate. And we build an analysis that says for each and every single one of those stores, what percentage or how much of that delivery could I have taken today based upon current state of technology? How much do I think I can take a year from now based upon the forecasted state of technology, two years from now, five years from now, et cetera.
And what that answers, and answers very concretely, is how well positioned are you, Company X, Y, Z, for the future of autonomy, whether it be with Nuro or with somebody else. Obviously we hope it’s with us. But that is a piece of technology and a piece of strategic thought partnership that I think is very, very difficult, if not impossible to get from someone else.
And part of what we do with that is we look at things that Google Maps or whatever, can’t look at, or doesn’t look at today. So we look at things like how many lanes there are in a road, what the speed limit of the road is, the slope of the road, the number of intersections, how complicated those intersections are, are they protected or unprotected? I think we even have the viscosity of the road in the model somehow.
So there’s all sorts of different factors that come into play that allow us to get very, very granular about the type of the deployments that we can do and where we can do them. And that then allows our partners to say, gosh, yes, this is worth investing in, or, no, that we’re not positioned. We shouldn’t continue to invest in this space because it’s never going to commercialize in a way that matters to us or the timeline for that is too far away.
Now, admittedly, you don’t tend to see that happen very often with restaurants, right. They tend to be in neighborhoods. They tend to be near dense populations. They tend to have volume. So restaurants by and large tend to do quite well in these things. But the tool itself is a very, very useful strategic tool for partners and for others to figure out how to make this succeed, and frankly, whether it will.
Cosimo, zooming out, how does Nuro help restaurants improve margins and upsell?
So there’s really two parts, right. So one is the core underlying simple math of you’re paying X for delivery today. Autonomous delivery’s going to be cheaper, right. And we think that’s already pretty significant. When you reduce the cost of delivery you’re going to see an explosion in delivery volumes, as we talked about. So that’s one part of it.
The other part though I think, which is secondary effect, or a second order effect, is what happens more broadly. So one of the things that’s not talked about a ton about delivery businesses is that you tend to see a much lower attach rate for high margin products like soft drinks or alcoholic beverages and things of that nature when they are delivered, as opposed to in-store. And that makes sense because you’re not typically thinking about, I want a Coke when you place an order for a pizza, but you definitely think about it when you’re eating that salty pizza and you get thirsty, right. So that’s not great for businesses.
Now, one of the things that we have the ability to do is change that paradigm. So instead of having to incent people to place those orders for Coca-Cola or whatever it might be at the moment of ordering, you can now actually incent them at the moment of delivery. And the way we would do that is the vehicle itself has two compartments, right. So imagine this, imagine one compartment has your hot pizza, your burrito, whatever it might be, and the other compartment has ice-cold high-margin products, right. They could be private label drinks. They could be Coca-Cola. They could be alcoholic delivery of some kind, all sorts of things of that nature.
In fact, by the way, a large number of these companies don’t do alcoholic delivery today because it’s just simply complicated and it’s fraught with challenges from a regulatory perspective. But imagine for instance that we deliver something to you and now, not only do you have the high-margin product being delivered at the moment of delivery, but you also have the opportunity to introduce products that you couldn’t introduce before, like beer, right.
So we have cameras, we have LIDAR, we have radar, you’re being recorded six ways of Sunday whenever we do a delivery. So it’s very, very easy for us to prove that we deliver to the right person, that we know that that person is of age, that they weren’t inebriated, right. We have the audit trails. All of those things are programmatically solved through autonomous delivery. And that enables a whole slew of new services and high-margin products that otherwise you could not do with human delivery or you just don’t do today because it’s just too burdensome.
That’s awesome, Cosimo. Thank you so much for that insight. Just my last question for you, unfortunately just a boring economics one. But curious how does Nuro make money from restaurants? Is there a fixed monthly lease or is payment variable?
So both, right. So what we want to do again is really build a solution that partners need. And so in some cases, right, some partners who have a lot of demand and who can take a full vehicle and utilize it well, it makes sense to have more of a fixed monthly lease, right. That enables certain things. That means that you can do things like co-branding on the vehicle. You can maybe get a little more creative with the inserts and do more fun things.
If you’re a brand or a restaurant that has less volume, right, where it probably doesn’t make sense to have a fully dedicated vehicle, in that case we would have much more of a just-in-time variable payment model whereby it’s perhaps not as accretive as the least model if you have a ton of volume, but it’s more flexible. And so it really depends on what the partner needs and how it fits best within their model.
And one last piece of that is, in some cases we have partners who are happy to have us park and charge at their facilities, and we have other partners for whom that is quite untenable because they just don’t have the real estate to do it. And so we’re able to play with all of these variables to come up with the right economic model so that we’re both aligned to ensure that we have the strongest chance of success together in launching, scaling the services.
Cosimo, this has really been a fascinating discussion today. Thank you so much for your time and insights and for keeping us posted on Nuro’s mission.
Awesome. Well, thank you guys. I really appreciate the time and I appreciate the invitation. It was an honor and privilege.
Thanks for joining us. Stay tuned for the next episode of Cowen Insights.
Get the Full Report
If you’re already a member of our Research portal, log in.Log In
If not, reach out to us directly for more information.